EP0360653A1 - Method and system for correcting image defects caused by the movement of a scanner - Google Patents

Abstract

Method for correcting in a scanner defects caused by movements of the scanner, which consists in disposing a test bar (P) which introduces a large attenuation and making the scanner perform a complete turn about a centre of rotation O and obtaining m distinct image captures of the test bar (P) each corresponding to an angular position alpha j. Analysis of the signals from each image capture enables determination of the angle beta j of the test bar, knowledge of this angle beta j leading to the calculation of the coordinates b and c of the test bar P and of the theoretical value beta th of the angle beta for each image capture. The difference between the theoretical value beta th and the measured value beta j is a correction value which is used in the device for formulating the scanner image. <IMAGE>

Description

The invention relates to X-ray scanners and more particularly to a method which makes it possible to correct, in the images obtained, the defects due to the displacements of the measuring assembly; it also relates to a system for correcting such faults.

To examine a patient, more and more x-ray machines called "scanners" are used which take images of the patient's cross sections. These devices are based on the physical phenomenon of absorption of X-rays by the human body. This absorption is directly linked to the distance traveled x of the X-rays in the body according to the formula:
I = I _o e ^-bx
formula in which:
I _o is the intensity of the radiation entering the human body
I is the intensity of the radiation leaving the human body,
b is a coefficient of attenuation which depends on the body crossed.
In a logarithmic measurement scale, the attenuation I / I _o is equal to bx, i.e. it is proportional to the distance x.

These devices consist essentially, as shown in Figure 1, of a source 10 of X-rays associated with a detection device 11, these two elements being arranged relative to each other in a fixed geometric relationship so to be able to interpose between them the body to be examined. In addition, they are supported by a structure (not shown) which can rotate around the body to be examined so as to irradiate the body at different angles. The X-ray source, which is controlled by a device 13, emits its rays in an angular sector which has a width sufficient to illuminate the entire cross section of the body. The detection device 11 has the form of an annular sector the length of which is adapted to the width of the X-ray beam and consists of a large number of elementary detectors 12 juxtaposed one next to the other.

To obtain an image of the cross section of the human body through which the X-ray beam passes, the support structure of the source 10 and of the detection device 11 is rotated around the body and the output signals of the elementary detectors 12 are measured. to treat them appropriately according to known methods in order to obtain a representative image of the cross section. For this processing, the elementary detectors 12, also called channels, are connected to an electronic device 14 which first performs a calculation of the logarithm of the signals received so as to obtain a signal whose amplitude is proportional to the attenuation of the rays X.

The axis of rotation of the structure - source 10, detection device 11 - is materialized by point 0, which means that the source 10 and the detectors 12 each describe a circumference of center 0 and of radius determined during rotation of the structure. This is only true in theory because the construction and use tolerances lead to deviations from the theoretical positions. This then results in defects in the images called "artifacts" and a decrease in spatial resolution.

The object of the present invention is therefore to implement a method which makes it possible to correct the defects resulting from deviations in the position of the source and of the detectors with respect to their theoretical position.

Another object of the invention is also to provide a system which implements said correction method.

The invention relates to a method for correcting image defects of a scanner due to the movements of the latter, characterized in that it comprises the following operations:
- The installation of at least one bar which introduces a significant weakening of the incident X-ray,
the rotation of the scanner over a complete revolution so as to produce a number m of views on the N detectors or channels of the scanner, each view corresponding to a determined angular position α _j of the scanner around its center of rotation,
- the determination, for each view, of the center of gravity of the attenuation (bar) so as to measure an angle β _j between the axes of radiation passing one by the center of rotation and the other by the center of gravity attenuation,
- the measurement of the Cartesian coordinates b and c of the center of gravity of the attenuation, that is to say of the bar,
- the calculation for each view of the theoretical angle β _th between the axes of radiation passing one by the center of rotation of the scanner and the other by the point of coordinates b and c,
- the calculation for each view of the difference δ _j between the angles β _th and β _j and the storage of the m values δ _j , each value δ _j then being used to modify the corresponding value of the position angular α _j .

The invention also relates to a system for implementing the method described above in a scanner comprising an X-ray source, an X-ray detection device with N detectors, means for rotating, around a center of rotation, the X-ray source and the detection device, means for introducing at least one bar between the X-radiation source and the X-ray detection device, means for recording in a memory the received signals corresponding to m different angular positions α _j of the source and of the detection device during a complete revolution, characterized in that it further comprises:
first means associated with the memory for analyzing the signals contained in said memory so as to determine for each view the value of the angle β _j ,
second means associated with the memory for analyzing the signals contained in said memory so as to determine the coordinates b, c of the bar,
- third means for calculating the theoretical angle β _th for each of the m angular positions α _j .
- fourth means for calculating for each of the m views, the angular difference
β _th - β _j = δ _j , and
- fifth means for recording the m angular correction values δ _j .

• - la figure 1 est un schéma de principe d'un scanner à rayons X ;
• - la figure 2 est un schéma géométrique qui permet de comprendre le procédé mis en oeuvre; et
• - la figure 3 est un schéma fonctionnel d'un système de traitement des signaux des détecteurs mettant en oeuvre le procédé de correction des images d'un scanner selon l'invention.

Other characteristics and advantages of the present invention will appear on reading the following description of a particular embodiment, said description being made with the accompanying drawings in which :

• - Figure 1 is a block diagram of an X-ray scanner;
• - Figure 2 is a geometric diagram which allows to understand the process used; and
• - Figure 3 is a block diagram of a signal processing system of the detectors implementing the image correction method of a scanner according to the invention.

The geometrical diagram of FIG. 2 makes it possible to better understand the process and to determine the corrections to be made. In this figure, the point O designates the axis of rotation of the scanner and the axes Ox and Oy are two axes orthogonal to each other and to the axis of rotation, one Ox being horizontal, the other Oy being vertical. The angular position of the structure is defined by the angel α made by the line FO with Ox, F materializing the focal point of the X-ray source. The point P indicates the location of a bar which strongly absorbs X-ray and whose coordinates are for example b for the abscissa and c for the ordinate.

Furthermore, the angle OFP will be called β; this angle β varies during a rotation cycle of the scanner. This angle can also be defined as being the angle between the central detector (aligned with FO) of the detection device 11 and the rank i detector whose signal is attenuated by the bar P. Thus, for each view of order j , several channels will present a signal strongly attenuated by the bar P, which will make it possible to define an angle β _j . According to the invention, this angle β _j will be compared by subtraction to the theoretical angle β _th defined by the trigonometric relation:

It is the result of this subtraction β _th - β _j which will be used to make corrections on the angular positions α _j .

To know β _th , it is first necessary to determine the coordinates b and c of the bar P, which can be obtained in different ways, for example by carrying out an acquisition on the bar leading to measuring b and c on the image obtained. Another way to calculate b and c will be explained below in relation to the description of Figure 3. The angle β _th is then then calculated by formula (1).

For each view j, the angle β _j is determined by the position of the channel i which presents the most attenuated signal, that corresponding to the X-rays meeting the bar P.

The method for correcting image defects of a scanner due to the movements of the latter therefore comprises the following operations:
- The installation of at least one bar introducing a significant weakening of the incident X-ray,
the rotation of the scanner over a complete revolution so as to produce a number m of views on the N detectors or channels, each view corresponds to a determined angular position α _j of the scanner around its center of rotation,
the determination, for each view of order j (angle α _j ), of the center of gravity of the attenuation (corresponding to the bar P) and thus obtain the value of the angle β _j ,
- the measurement of the coordinates b and c of the bar P,
- the calculation for each view j of the theoretical angle β _th from the measured coordinates b and c of the bar using the formula (1);
- the calculation for each view j of the difference
δ _j = (β _th - β _j )
which constitutes a correction value and the storage of the m values δ _j ;
- the use of each value δ _j to modify the corresponding value of the angular positions of the N detectors for the view of order j.

To implement this correction method, the invention proposes a system which will now be described in relation to FIG. 3. It comprises an analog / digital converter 20 to which the output signals of the N detectors or channels of the detection device are applied. 11. The N digital codes corresponding to a position or view of rank j are applied to a logarithm calculation circuit 21 which provides, for each channel i and each view j, a code representative of the attenuation undergone by the X-rays. N codes resulting from this logarithmic calculation operation are applied to a subtraction circuit 22 in which a REF value representative of the attenuation undergone by the X-ray in the air is subtracted from them. This value is obtained using a detector called a monitor, the position of which on the detection device 11 is such that it receives X-radiation without attenuation.

The codes which result from this subtraction, called measurements Y _ij , are recorded in a memory 29 then processed in a calculation device 23 in order to determine for each view j, the angle β _j corresponding to the bar P. This is for example carried out by analyzing the N signals Y _i of each view to determine the channel i which represents the center of gravity of the attenuation due to the bar P. More precisely, the channels which correspond to signals attenuated by the bar P are determined, attenuation of which the maximum does not necessarily correspond to the exact angular position of a channel . In other words, the straight line FP is not generally aligned with the center of a detector 12 (Figure 2). Also, it is necessary to determine the maximum of the attenuation and one then obtains a value of i which is not integer but fractional. The angle β _j is then calculated by the number of channels or detectors 12 which are between the central detector, aligned with FO, and the detector of rank i (i fractional) corresponding to the maximum attenuation, taking into account the fact that the detectors are regularly arranged on an arc of a circle of center F.

The m values β _j are recorded in a memory 24 in order to serve, on the one hand, as a basis for calculating the coordinates b and c and, on the other hand, to obtain the correction values δ _j .

avec j variant de 1 à m de manière à obtenir m/₂ systèmes d'équations. On comprendra que β_m+1 correspond en fait à β₁.The calculation of b and c is carried out in a circuit 25 using the formula (1) in which β _th is replaced by β _j measured. More precisely, the successive values of β _{j are used} two by two so as to obtain each time a system of two equations with two unknowns b and c in which β _j and α _j are known. Each system of equations is of the type:

with j varying from 1 to m so as to obtain m / ₂ systems of equations. It will be understood that β _{m + 1} in fact corresponds to β₁.

The resolution of the m / 2 systems results in m / 2 values of b and m / 2 values of c whose averages are calculated.

These are the average values of b and c which are transmitted to a circuit 26 which performs for each position view α _j the calculation of the m theoretical values β _th .

Each of the m values β _j contained in the memory 24 is subtracted algebraically in a circuit 27 from the corresponding value (same angle α _j, that is to say same view) which is provided by the calculation circuit 26. We then obtain m differential values δ _j which are recorded in a memory 28 for use in the image processing device by modifying the angles α _ij which indicate the angular positions of the detectors with respect to the axis Ox.

The invention has been described using a single bar, but it can also be implemented using several bars at different positions.

The system which has been described in relation to FIG. 3 can be produced according to different variants without departing from the scope of the present invention. In particular, the memory 29, which records the values Y _ij , can be deleted if the calculation of β _j (circuit 23) is carried out in real time.

Likewise, the calculation of the coordinates b and c of the bar P can be carried out in different ways, for example on the image of the bar obtained by the scanner as indicated above.

Claims (5)

1. Method for correcting image defects of a scanner due to the movements of the latter, characterized in that it comprises the following operations:
- The installation of at least one bar (P) which introduces a significant weakening of the incident X-ray,
the rotation of the scanner over a complete revolution so as to produce a number m of views on the N detectors or channels of the scanner, each view corresponding to a determined angular position α _j of the scanner around its center of rotation,
the determination, for each view, of the center of gravity of the attenuation so as to measure an angle β _j between the axes of radiation passing one by the center of rotation and the other by the detector having the center of gravity of mitigation.
- the measurement of the Cartesian coordinates b and c of the center of gravity of the bar (P),
- the calculation for each view of the theoretical angle β _th between the axes of radiation passing one by the center of rotation of the scanner and the other by the point of coordinates b and c,
- the calculation for each view of the difference δ _j between the angles β _th and β _j and the storage of the m values δ _j , each of these m values then being used to modify the corresponding value of the angular position α _j . 2. Method according to claim characterized in that the operation of determining the angle β _j consists in analyzing, for each view, the N signals received from so as to determine the angular position of the center of gravity of the weakest signal and to compare it to the angular position corresponding to the focal point alignment (F) of the X-ray and of the center O of rotation of the scanner. 3. Method according to claim 1 or 2, characterized in that the measurement operation of the Cartesian coordinates b and c of the bar (P) is carried out from the m values β _j by solving m / 2 systems of two equations with two unknown type:

(j varying from 1 to m) and by calculating the means of the m / 2 values obtained for b and c. 4. Method according to claim 1, 2 or 3, characterized in that the calculation operation for each view of the theoretical angle β _th is carried out according to the formula:

5. System for implementing the method according to any one of the preceding claims in a scanner comprising a source (10) of X-radiation, a device (11) for detecting X-radiation at N detectors, means for rotating the X-ray source and the detection device around a center of rotation, means for recording in a memory the signals received corresponding to m different angular positions α _j of the source and the detection device during a complete revolution, characterized in that it further comprises:
- first means (23) associated with the first memory for analyzing the signals contained in said memory so as to determine for each view the value of the angle β _j ,
- second means (25) for determining the coordinates b, c of the bar (P),
- third means (26) for calculating the theoretical angle β _th for each of the m angular positions α _j .
- fourth means (27) for calculating for each of the m views, the angular difference
β _th - β _j = δ _j , and
- fifth means (28) for recording the m angular correction values δ _j .

Images